Special Interest Group Report Out
- Eric Jansson
The SIG set out to accomplish three analytical objectives in the area of understanding the Ed-Fi Community's needs for open source and alternative database platforms for the Ed-Fi technology artifacts, and especially the ODS/API. References to these original questions can be found in the SIG overview and in meeting notes.
The results of these are listed below.
1.Review and prioritize community needs and use cases in this area
The main concern of the Ed-Fi community is that the cost of SQL Server licensing is high, especially in cloud deployments where existing organizational licensing may not apply. Examples of costs estimates where the licensing costs were estimates to be 6x to 10x the costs of operational fees for service-based offering were provided (note that these are slightly different cost types). In addition to this core problem, the following were related points or conclusions raised by the SIG:
- The costs of scaling linear-ally (such as in LEA collaboratives where each LEA has a physically separate database or database platform) were noted to exacerbate the high costs of licensing
- System integrators and others in similar roles noted that they wanted more ability to offer customer choice of platforms
- The ability to develop solutions reusable across the K12 sector, and to not have local licensing as a barrier to such adoption
- "Philosophical" opposition to proprietary software and/or preference for open source as a community good was cited several times
2. Provide actionable recommendations for the Ed-Fi Technology Roadmap specifically
- Add support for a commonly used, open source relational database alternative. The SIG felt the strongest choice was likely PostgreSQL.
- The SIG felt that this support would likely apply transitively to cloud-service based offerings that were based on the PostgreSQL interfaces, so no additional roadmap recommendations for that were made by the SIG.
- At this time, do not prioritize work looking into NoSQL/unstructured/document store type databases. These products likely have their place and usage, but scale and need in the current Ed-Fi ecosystem is addressable through relational systems.
- The SIG provided no clear direction on 2.x vs 3.x support, and noted the value of both (encouraging upgrade vs meeting the community where it is)
3. Provide recommendations for the Ed-Fi Community member work in this area more broadly
This question was not tackled explicitly, as the main advice was to add support for a single open source alternative and so other community work was not envisioned.